Page 6 of 11
Appendix Table 3. Case definitions for variants of concern*
Case definition (* denotes any number indicating a sub-lineage)†
SARS-CoV-2-positive cases whose samples were confirmed through whole-genome sequencing as Pangolin lineage
B.1.617.2 or AY.*
SARS-CoV-2-positive cases whose samples were confirmed through whole-genome sequencing as Pangolin lineage
B.1.1.7 or Q.*, or screened as N501Y-positive and E484K-negative
SARS-CoV-2-positive cases whose samples were confirmed through whole-genome sequencing as Pangolin lineage
P.1 or P.1.*, screened as N501Y- and E484K-positive, or screened as K417T-positive
SARS-CoV-2-positive cases whose samples were confirmed through whole-genome sequencing as Pangolin lineage
B.1.351 or B.1.351.*
*VOC, variant of concern.
†Screening employed targeted single-nucleotide polymorphism quantitative polymerase chain reaction (SNP qPCR).
‡Based on epidemiology at the time, VOC screening results with both E484K and N501Y mutations were assumed to be Gamma VOC, given the
very low prevalence of Beta VOC in BC.
Appendix Table 4. Modeling approach and parameters used for each counterfactual scenario
Parameters and explanation
Baseline transmission was fit using a Bayesian
approach to daily BC PCR-confirmed COVID-19
cases. Transmission was modeled as a piecewise
constant function with pre-specified breakpoints.
Breakpoints were selected based on changes in public
health measures or population behavior. The start dates
for each segment were: 2020–03–15, 2020–05–10,
2020–09–08, 2020–09–21, 2020–11–10, 2021–01–25,
2021–03–29, 2021–04–05, 2021–05–25.
Delta variant
Logistic growth curve multiplied by transmission
rate β to replicate the relative increase in
transmission as the Delta variant became dominant.
The parameters include the time of introduction and
the time to the proportion of Delta being 90%
among all variants.
The BC scenario ramp had a start date of 2021–05–25
and a time to dominance of 25 weeks. The England
scenario had a start date of 2021–05–01 and a time to
dominance of 4 weeks to reflect the differences in the
proportional change in Delta between both jurisdictions.
For each scenario, Delta increased transmission by 50%
measures and
health behavior
A further change in the transmission was included
to characterize the changes in public health
measures and behavior following their introduction
on 2021–04–05 in BC. This was modeled as a step-
change in the transmission term after
implementation of the measures.
In the BC scenario, the transmission term was fitted to
historical data and so no further changes needed to be
included. For the England scenario, the transmission
term is multiplied by a factor of 50% reflecting no
changes in public health measures following 2021–04–
out and coverage
The exact proportion of vaccination by age group
were extracted for both BC and England
jurisdictions. For each vaccination coverage, an
estimate of BC’s population structure and contact
rates by age group were used to derive an age-
adjusted vaccination rate for each scenario.
Proportion of each age group vaccinated based on data
for BC and England. Number of weekly contacts per age
group based on BC-Mix survey data (18) were:
<2 y: 6, 2–5 y: 12, 6–17 y: 13, 18–24 y: 44, 25–34 y: 44,
35–44 y: 48, 45–54 y: 48, 55–64 y: 24, 65–74 y: 21, >75
product
The vaccine transmission-blocking effectiveness
was selected for each scenario based on the
dominant vaccine product for each jurisdiction
during the study period. In BC this was BNT162b2
and in England this was ChAdOX1.
Vaccine effectiveness for the BC scenario was 71.8%
after one dose and 89% after two doses. Vaccine
effectiveness for the England scenario was 58% after
one dose and 77% after two doses.
Appendix Table 5. Transmission-blocking efficacy parameters used in counterfactual modeling by vaccine product and first and
second dose
*Efficacy parameters for vaccination incorporated both a reduction in the
probability of incidence infection as well as reduction in onward
transmission if infected while vaccinated. Estimates for vaccine
transmission-blocking efficacy and onward-transmission efficacy were
derived from values presented in the UK Scientific Pandemic Influenza
Group on Modeling, Operational subgroup (SPI-M-O) of the Scientific
Advisory Group for Emergencies (SAGE) July 7th modeling group report
(19) where a range of optimistic to more pessimistic vaccine assumptions
were provided by different modeling groups. The total transmission
blocking probability is 1-(1–0.6)*(1–0.45) = 0.78 under the optimistic
scenario and 1-(1–0.53)*(1–0.4) = 0.718 under the pessimistic scenario for
BNT162b2, rising to 0.89 under the second dose. For ChAdOx1, after one
dose, this is 0.58 under the pessimistic and 0.65 under the optimistic
scenarios, rising to 0.77 under the second dose pessimistic and 0.8 under
the optimistic scenarios, respectively. Scenarios in British Columbia were
more in line with pessimistic assumptions for BNT162b2, so for ChAdOx1
we also selected the pessimistic assumption.